Published on: 2020-05-17T09:11:33+00:00
The recent discussion on the bitcoin-dev mailing list focused on several key topics related to the Lightning Network (LN) and the importance of maintaining a decentralized network of full nodes.One major concern raised was the need to make it easy for mobile LN clients to connect to home full nodes, in order to prevent centralized services from gaining an advantage. Christopher Allen introduced QuickConnect as a potential solution, which allows for secure connection via Tor v3 from a remote device to a personal full node. However, improvements are still needed for QuickConnect, and contributions for the next version are welcomed.There were also discussions about the economic weight of nodes in evaluating miner consensus-hijack success and the hope that Lightning will replace centralized custodial services. The scalability advantage of Lightning lies in its ability to transfer without touching the blockchain, resulting in reduced fees while maintaining noncustodiality. However, concerns were raised about the distribution of economic weight in a Lightning future with massive Simplified Payment Verification (SPV) nodes. It was suggested that backup nodes should be picked up manually at client initialization to avoid conflicts during contentious periods or social engineering attacks. Incentivizing publicly-facing backup servers and privately-owned servers were proposed as alternative solutions.To ensure a seamless user experience, suggestions were made to improve the installation and configuration process of server applications and to educate users on differentiating between client and server apps. "Zero configuration" apps or easy in-app walkthroughs were mentioned as possible solutions. It was also noted that once a full node has synced the whole blockchain, only the header chain is necessary, reducing the need to carry 100% of the block data. However, trust in the node to which Lightning is tethered remains a concern.Antoine Riard and Chris Belcher discussed the impact of Lightning Network on the trust-minimization of Bitcoin's security model. Antoine suggested that LN's fast, affordable, confidential, and censorship-resistant payment services may attract adoption without users running a full node. Chris argued that this compromises Bitcoin's security, as it relies on the assumption that a supermajority of the economy verifies incoming transactions using their own full node. Failure to do so could result in Bitcoin splitting into two currencies: full-node-coin and SPV-coin. Antoine proposed considering the economic weight of nodes in evaluating miner consensus-hijack success.The discussion also touched on the challenges of relying on local network peers in LN and the potential risks of collusion among direct LN peers to not forward payments, rendering the money valueless. Various solutions were proposed, including alternative HTLC constructs, public watchtowers, HTTP proxy data cache, and incentivizing full nodes to exchange new data. Privacy concerns and the need for uniform distribution of backup nodes were also addressed.ZmnSCPxj highlighted the limitations of opportunistic payment channels for paying for validation information in LN and suggested an alternative payment system similar to watchtowers. Christopher Allen expressed interest in hosting a collaboration for defining different sets of wallet functionality and suggested creating a transparent shim between Bitcoin-core and remote RPC with cryptographic capability mechanisms to control access to specific node functions. Refactoring in Core wallets is making progress, with feedback and proposals for usability improvements.The conversation also discussed the feasibility of running full nodes on everyday phones and the challenges in designing a mobile-first LN experience. The idea of part-time or Sleeper Nodes™ was proposed, where phone nodes store and serve a randomly selected range of blocks during nighttime operation. Concerns about scalability, security, and privacy of LN for millions of clients were raised, as well as the need to incentivize node operators to dedicate resources. Monetary compensation for servicing filters was suggested as a solution.There were discussions about improving interfaces for serving peer services and implementing BIP 157, which improves privacy for full-node-less usage. Suggestions were made to support light clients and encourage full node diversity through alternative last-mile networks like geostationary satellites or radio systems. The shift in the security model of Bitcoin as more light clients outpace full nodes was a major concern, and the importance of allowing light clients to explicitly choose their full node tethers was emphasized.In conclusion, the email exchange discussed various challenges and potential solutions related to connecting mobile LN clients to home full nodes, the role of economic weight in nodes, the risks of relying on local network peers, and the need for alternative payment systems and backup nodes. The discussion also highlighted the importance of maintaining a decentralized network of full nodes and the potential risks of compromising Bitcoin's security by promoting a "full node-less" experience.
Updated on: 2023-07-31T22:55:16.399575+00:00